Estimating Construction and Demolition Waste Through Automatic Detection of Building Changes: A Building Inventory Approach

The escalation of urban development and infrastructure projects has necessitated a comprehensive understanding of construction and demolition (C&D) waste volumes, which pose significant environmental challenges. This study presents a novel framework designed to estimate C&D waste volume through an integrated approach that combines the automatic detection of building changes in the inventory and field assessments of building material compositions in Wallonia, Belgium. By leveraging geospatial data from over two million buildings, we classify them into four groups: demolished, new, modified, and unmodified. Initially, we have identified 37,317 demolished buildings, a figure that is comparable to the demolition records available over the past ten years. Subsequently, we employ a logistic regression model to analyze the relationship among building characteristics, land parcels, and patterns of building demolition. The findings reveal critical insights into factors influencing C&D waste volume. Furthermore, we set up a sampling strategy for analyzing the composition of residential building materials, with a focus on calculating material intensities. The material intensities are cross-referenced with the gross volumes of the buildings to obtain an estimate of the materials present in the residential building stock. Based on the validated demolition rate, we estimate the volume of waste generated from the demolition of buildings that can be recycled or reused. Finally, we evaluate the derived C&D waste using the data from the recycling and demolition centers in Wallonia.